Differentiating between the interactions available for task FILL

The following question is addressed in this section: if you want to construct a test question which presents your candidates with a FILL task, how do you decide between the interactions available? 

The GUIDANCE differentiator

The main difference between these three interactions when used for a FILL task is probably the level of guidance (if any) given to candidates, and thus the level of difficulty of the test question. The big difference between the Text Entry interaction and the other two is that candidates receive no hints about what the answer is. In contrast, both Inline Choice and Gap Match provide a list of options which candidates choose from. The number of choices provided varies between these two latter interactions, however:  if the test question contains a long text with a large number of gaps, candidates have more options for each gap if a Gap Match interaction is used. With an Inline Choice interaction, the number of choices offered is more easily - and probably more often - limited in the drop-down menus for each gap. So, the factor which will probably most influence your choice is the level of guidance you want to give candidates. 

When we plot the three interaction types on the GUIDANCE spectrum as follows:

We can see that Inline Choice generally offers slightly more guidance than Gap Match, whilst Text Entry offers none at all. This gives us an indicator when choosing between the three main interactions for a particular test question of type FILL.

SCOPE as a factor

There is another factor, however, which is relevant to your choice of interaction, and this is the scope of the test question. Inline Choice interactions give you the possibility of offering words or phrases in the drop-down menus which aren’t in the text at all, thus opening up more options for what you can test than a Gap Match interaction might. You might want to test verb endings, for example, and therefore to offer as potential responses the same verb with different endings. It's not impossible to limit the scope of the question in such a way with either Gap Match or Text Entry interactions, but using an Inline Choice interaction is the easiest and most obvious way to do this.

To sum up, the level of guidance you wish to offer test candidates will be an important indicator in your choice between the three interactions available for FILL tasks, but it will also be useful to bear in mind what type of information you are testing, and that it's easier to define specific possible responses - if you want to assess a particular phenomenon - with an Inline Choice interaction.